1854
products. For purposes of following spectrophotometrically the changes in product structure occurring during the course of reaction, 1 : 6 vjv trifluoroacetic acid-carbon tetrachloride was used as the medium in this phase of the investigation. In solutions in this mixture in which the original concentrations of o-nitroM , isoxazole (11) benzhydryl bromide were 2-4 X and nitroso ketone (I) were observed to form simultaneously, the latter forming rapidly during early phases of reaction and then much more slowly as hydrogen bromide in the medium increased. After at least 10 half-lives elapsed, the yields of isoxazole and nitroso ketone were of the order of 70-80z and 15-20z. In 0.1 M concentration in acetic acid the o-nitro compound solvolyzed in the initial phases of reaction to produce a which decreased small amount of nitroso ketone (6-7 in quantity as reaction proceeded due to formation of isoxazole as the exclusive final organic product. The conversion of nitroso ketone to isoxazole by hydrogen bromide, which is very rapid in acetic acid solutions of 0.01-0.1 M hydrogen bromide concentration, is apparently somewhat slower in trifluoroacetic acid. In acetic acid in the presence of sodium acetate, o-nitrobenzhydryl bromide solvolyzes to produce exclusively o-nitrosobenzophenone. The sodium salt in
z),
this case prevents accumulation of hydrogen bromide in the medium. In 1 :6 vjv trifluoroacetic acid in the presence of sodium trifluoroacetate the formation of nitroso ketone from the halide is not blocked, and in fact the nitroso ketone and isoxazole are formed in about the same proportions as when salt is omitted from the reaction mixture. It appears that trifluoroacetate ion is a weaker base than bromide ion in trifluoroacetic acid. l7 In this same context it is interesting to note that nitroso ketone formation in the reaction of the o-nitro compound in trifluoroacetic acid-carbon tetrachloride can be eliminated completely by adding either hydrogen bromide or lithium bromide. Rate constants, based on the measurement of formation of isoxazole as the reaction product in solutions initially 2-4 X M in Meither in o-nitrobenzhydryl bromide and 6-10 X hydrogen bromide or lithium bromide, were found to be in good agreement (kz60 9 X sec-l). Eaborn, Jackson, and Taylor have observed previously that hydrogen bromide is evolved from a solution of sodium bromide in trifluoroacetic acid. l8
-
(17) Cf.J. Bessiere, Bull. SOC.Chim. Fr., 9,3353 (1969). (18) C. Eaborn, P. M. Jackson, and R. Taylor, J. Chem. Soc. E, 613 (1966).
Aminolysis Reactions. 11. Catalysis of Ester Aminolysis in Chlorobenzene. Correlation with Hydrogen-Bonding Ability of Catalysts Chih-Wu Su and Joseph W. Watson* Contributionf r o m the Department of Chemistry, The Third College, University of California, San Diego, La Jolla, California 92037. ReceivedNovember 12,1973 Abstract: The n-butylaminolysis of p-nitrophenyl acetate in chlorobenzene at 25' follows the rate equation kl.obsd = k3(BA)
+
k3'(BA)(cat.). Dipolar oxygen bases are very effectivecatalysts of the reaction; N,N-dimethylacetamide is more than twice as effective in catalyzing the reaction as n-butylamine although it is a 1O1O times weaker aqueous base. The catalytic abilities of several classes of oxygen and nitrogen compounds correlate with their hydrogenbonding ability to p-fluorophenol in chlorobenzene by the equation log kg' = 1.2 log Kr - 3.3. This correlation indicates that there is 10-30x proton transfer to the catalyst in the transition state. The results suggest that amides could serve as general base catalysts in enzymes with hydrophobic active sites.
I
n 1969 we1 presented evidence and arguments against the bifunctional or cyclic nonpolar mechanisms for the aminolysis and amidinolysis of p-nitrophenyl acetate, p-NPA, in chlorobenzene solvent. We proposed that the reactions were general base catalyzed and heterolytic in mechanism and suggested the use of unhindered tertiary amines, such as lY4-diazabicyclo[2.2.2]octane (DABCO), and intramolecular tertiary amine catalysts as mechanistic probes to differentiate between cyclic and stepwise general base catalysis. (1) H. Anderson, C. Su, and J. W. Watson, J. Amer. Chem. Soc., 91, 482 (1969). (2) (a) F. M. Menger, J. Amer. Chem. Soc., 88, 3081 (1966); (b) p. L. Lillford and D. P. N. Satchell, Chem. Ind. (London), 1750 (1967); J. Chem. SOC.B, 360 (1967); (c) S.T. McDowell and C. J. M. Stirling, ibid., 343 (1967); (d) A. Shawali and S. S. Biechler, J . Amer. Chem. Soc., 89,3020 (1967).
Journal of the American Chemical Society 1 96:6
Subsequent work by Pietra, et al., has presented additional evidence in favor of the bifunctional catalysis mechanism for benzamidinolysis reactions. 3a Although additional arguments have also been presented on behalf of the bifunctional catalysis mechanism for ester a m i n o l y ~ i s ,the ~ ~ most ' ~ recent work of Menger, et a]., provides compelling evidence that the aminolysis of aryl esters in aprotic solvents proceeds by a general base-catalyzed mechanism of the general form given in eq l.4 It is surmised that the base catalyst, which may be a second molecule of the nucleophilic amine, func(3) (a) G. Biggi, F. Del Cima, and F. Pietra, J . Chem. Soc., Perkin Trans 2, 188 (1972); (b) F. M. Menger and J. H. Smith, J. Amer. Chem. Soc., 91, 5346 (1969); (c) D. P. N. Satchell and I. I. Secemski, J. Chem. SOC. B, 1013(1970). (4) (a) F. M. Menger and J. H. Smith, J. Amer. Chem. Soc., 94, 3824 (1972); (b) F. M. Menger and A. C. Vitale, ibid., 95,4931 (1973).
March 20, I974
1855
Table I. Rate and Physical Constants of Catalyzed n-Butylaminolysis ofp-NPA in Chlorobenzene at 25” ks
Catalyst
[Catalyst]
1. None 2. N,N-Dimethyl-4-aminopyridine 3. DABCO 4. 1-Methylimidazole 5. 2,6-Dimethylpyridine 6. 4-Methylpyridine 7. Pyridine 8. Triphenylarsine oxide 9. Pyridine N-oxide 10. 2-Pyridone 11. 1-Methyl-2-pyridone 12. N,N-Dimethylacetamide 13. Trimethylphosphine oxide 14. Tetrahydrofuran 15. Triphenylphosphine oxide 16. Dimethyl sulfoxide 17. Benzonitrile 18. Sulfolane
0 . 0 0 5 4 .01 0.05-0.1 0.00124.061 0.018-0.063 0.01-0.061 0.06-0.084 0.0014-O.0035 0.001-0.0054 0.0021-0.14 0.005-O.011 0.014.052 0 . 0 0 2 3 4 .0079 0.0514.14 0.00214.0061 O.OO5-O).013 0.01-0.059 0.0033-0.017
x
H“
Db
M - 2 sec-l
k,’ X 10’ M-’ sec-1
3 + log k 3 ’
10.66e 9 . 70e
1.76 2.4
1.4 4.31
6.2d 5 . 6 10.2
32i 2
1 .79c 2.51
8.6f 7.2e 6.60e 6.03e 5.21e 0.99h 0. 79i 0.75; 0 . 32k -0.39m -2” -2.08P -2.10h -2.78” -10.79 -12.8’
1.84 2.41 1.78 1.69 1.55 3.17 2.35
3.6 1.66 2.60 2.20 5.54 4.20 2.10 4.00‘ 3.81 4.29O 1.75 4.31 3.9 4.05 4.69
6.49 8 . 3 11 . 4 6.3 f 0.4 5.9 f0 . 2 7 . 2 i 0.4 6 . 3 10 . 4 6.4 i 0.4 5.7 5 . 9 10 . 5 6.61. 0.3 5.5 & 0.5 6 . 3 i0 . 3 6.3 1. 0 . 3 6 . 3 1 0.5 5 . 5 i0 . 2 5 . 9 i 0.4
p&mi
2.00 2.00 3.02 0.99 2.72 2.14 0.56 1.21
102
8.410.1g 2212 0 . 9 + 0.7 61.1 2.1 1 0 . 5 4ooi 0 41 f 8 1OOoi 100 20 & 6 15.3 =t0 . 2 1 9 0 i 10 0.41.0.3 57i4 14+ 3 3 1 2 2* 1
1.92 2.34 0.95 1.78 1.3 3.60 2.6 4.0 2.3 2.18 3.25 0.6 2.76 2.28 1.5 1.3
~~
Dipole moments: A. C. McCellan, H values in chlorobenzene calculated from H values in CCl, using constants of Table 111, ref 7c. 3 log k Sof u-butylamine. Reference 2a. “Tables of Experimental Dipole Moments,” W. H. Freeman, San Francisco, Calif., 1963. J J. Hine, J. G. Houston, e D. D. Perrin, “Dissociation Constants of Organic Bases in Aqueous Solution,” Butterworths, London, 1965. C. Klofntar, F. Krasonec, and M. Kusar, Croat. Chem. J. H. Jensen, and J. Mulders,J. Amer. Chem. SOC.,87,5050(1965). 0 Reference 1. H. E. Zaugg, J. Amer. Chem. SOC.,82,2903 (1960). 7 S . F. Mason, J. Chem. Soc., 674 (1958). A. Albert and J. N. Acta, 40, 23 (1968). P. Haake, R. D. Cook, and G. H. H. Lumbroso a n d D . Bertin, B U N . SOC.Chim. Fr., 1728 (1970). Phillips, J. Chem. SOC.,1294(1956). R. S . Armstrong, M. J. Aroney, P. Haake and R. D. Cook, Tetrahedron Lett., 427 (1968). Hurst, J , Amer. Chem. SOC.,89, 2650 (1967). F. Le, J. W. Raymond, R. K. Pierens, J. D. Jaxby, and C. J. Wilkins, J. Chem. SOC.A , 2735 (1969). p E. M. Arnett, Progr. Phys. Org. S . K. Hall and E. A. Robinson, Chem., 1,325 (1963). q N. C. Deno, R. W. Gaugler, and M. J. Wisotsky, J. Org. Chem., 31,1967 (1966). Can.J. Chem.,42,1113 (1964).
+
II
KC‘NK
+
0.AF
tions to remove the ammonium proton of the tetrahedral intermediate and thereby enables decomposition of the intermediate to the neutral amide rather than to the much more energetic N-protonated amide.4b We report in this paper (1) that a variety of oxygen and nitrogen bases serve to catalyze the n-butylaminolysis ofp-NPA, (2) that the catalytic abilities of the oxygen bases are “exceptionally enhanced” relative to those of the nitrogen bases in this reaction, and (3) that the catalytic abilities of all the bases investigated with the exception of 2-pyridone which functions as a bifunctional catalyst5 are correlated with their hydrogen-bonding abilities rather than their aqueous basicities. The mechanistic implications of these findings are commented on. Preliminary reports of this work have been published.6 Results and Discussion In the presence of catalysts, the n-butylaminolysis of p-NPA in chlorobenzene follows the rate expression 1 , 2 kl ,obsd = kz(amine)
+ k3(amine)z + ka’(amine)(cat.)
The kz term was not detectable at 25” but is observable at higher temperatures.6b With secondary amines kz terms are observed at 25°.4$6b ( 5 ) P.R. R o w . J . Amer. Chem. Soc.. 91.6090 (1969).
(6) (a) P. W. Arana, C. Su, and J. W. Watson,’ Chem. Commun., 363 (1970); (b) C. Su, Doctoral Dissertation, University of California, San Diego, 1972.
The kS‘ values listed in Table I represent the average of several kinetic runs covering at least two concentrations of catalysts and at least three concentrations of n-butylamine for each catalyst’s concentration. A typical data plot is shown in Figure 1 for N,N-dimethylacetamide. The concentrations of catalysts were kept as low as possible to minimize medium effects. The relative constancy of the k, terms around its value in the absence of catalysts of 0.062 M - 2 sec-l suggests that medium effects due to the presence of the catalysts are small. If the k,’ values were due to medium effects, catalysts with high dipole moments such as sulfolane and benzonitrile would be expected to exhibit large medium effects in the low-polarity, low-dielectric constant solvent chlorobenzene and consequently have large k3’ terms. This is not the case. Sulfolane and benzonitrile have very small kg ’ terms suggesting that medium effects are small. Thus, we consider the k8’ terms to represent a specific catalysis effect and not a medium or solvent effect. In addition to listing the k3, k3’, dipole moments D, and pK, for each catalyst, Table I lists the H values for the catalysts. The H values are a measure of the abilities of the catalysts to form hydrogen-bonded complexes with p-fluorophenol in chlorobenzene ~ o l v e n t . ~
+ B +P-FPhOH. lil
p-FPhOH
log Ki
.B
H
With a single class of catalysts and in the absence of severe steric hindrance, the data of Table I indicate that catalytic ability follows aqueous basicity with a shallow slope; for the pyridines A log k3‘/ApK, = 0.3. How(7) (a) D. Gurka and R. W. Taft, J. Amer. G e m . Soc., 91. 4794 (1969); (b) R. W. Taft, D. Gurka, L. Joris, P. Schleyer, and J. W. Rakshys, ibid., 91, 4801 (1969); (c) L. Joris, J. Mitsky, and R. W. Taft, ibid., 94, 3438 (1972).
Su, Watson
1 Catalysis of Ester Aminolysis in
Chlorobenzene
1856
1.0
3.0
2.0
4.0
H
Figure 2. Plot of log k3' for cataiyzed n-butylaminolysis of p-NPA us. Hvaluein chlorobenzene at 25".
/ /
/
ever, across classes of catalysts there is no correlation of catalytic ability with aqueous basicity. The k3' values indicate that the dipolar oxygen bases are exceptionally effective catalysts exhibiting catalytic abilities equivalent to those of nitrogen bases which are more than lo7 times stronger bases in water ; N,N-dimethylacetamide is twice as effective a catalyst for the n-butylaminolysis of p-NPA as DABCO, although its aqueous pK, is 9 pK, units smaller (pK, -0.39 us. 8.6). Figure 2 indicates that the catalytic abilities are correlated with the hydrogen-bonding abilities of the catalysts by the equation log k,' = 1.2H - 3.3. Since AHjApK, for several classes of compounds is 0.21ib and the slope of Figure 2 is 1.2, A log k3'/ApK, for all classes of catalysts should be approximately 0.25, which is close to the value of 0.3 estimated from the pyridine data of Table I. The inclusion of the catalytic abilities of n-butylamine and DABCO to catalyze the n-butylaminolysis of p-NPA in the correlation provides strong evidence against proposals that primary and tertiary amines The catalyze the reaction by different mechanisms. possibility that the tertiary amines and oxygen bases catalyze the reaction by nucleophilic catalysis3cis made unlikely by the shallow dependence of catalytic ability on basicity, A log k,'/ApK, = 0.3. In terms of the mechanism of eq 1, the correlation of log k3' with H suggests that the catalysts effectively facilitate the expulsion of the phenolate anion from the tetrahedral intermediate by acceptance of the ammonium proton in the transition state to an extent equivalent to that in hydrogen-bonded complex formation with 3 b 3 c
p-fluorophenol. It is estimated that in such hydrogenbonded complexes the extent of proton transfer is 1030 of full proton The effectiveness of this small extent of proton transfer in facilitating the decomposition of the tetrahedral intermediate to products indicates that solvents capable of accepting hydrogen bonds (e.g., ethers or amides) can intervene and catalyze the aminolysis of aryl esters and that nucleophiles with greater internal abilities to stabilize positive charges than primary amines (e.g., secondary and tertiary amines and amidines) may be capable of expelling the phenolate anion from the tetrahedral intermediate without the necessity of proton transfer to a general base. There has been interest in the possible mechanisms of action of enzymes with hydrophobic active sites. Our results indicate that with such enzymes attention should be given to the possibility that dipolar groups generally not considered to be general bases may function as general base catalysts, especially amides and other dipolar oxygen compounds. For example, in this study the amide N,N-dimethylacetamide is essentially as effective a catalyst of the ester aminolysis as the imidazole N-methylimidazole and more effective a catalyst than the primary amine n-butylamine, although it is respectively a loi to 1Olo times weaker aqueous base. Thus, we have the intriguing prospect that an amide group of an enzyme with a hydrophobic active site may function as a general base catalyst. A subsequent paper will present data which suggest that the relative kinetic "basicities" of different classes of bases vary with the extent of proton transfer in the reaction. lo 2 , 4 b 3 9
Experimental Section All distillations and preparations of /?-butylamine,DABCO, and chlorobenzene were conducted as previously described. * Reagent grade pyridines were twice distilled from either KOH-zinc dust or calcium hydride under nitrogen. Gas chromatography on 5 ft Carbowax or XF-1150 fluorosilicone columns showed only single peaks. (8) D. Gurka, R. W. Taft, L. Joris, and P. Schleyer, J . Amer. Chem. SOC.,89, 5957 (1967). (9) R. L. Snell, W. I
